Conclusions:

The objective of the study was to undertake a detailed flood study of the Main Drain J catchment and establish models as necessary for design flood level prediction. This included the assessment of inputs to the catchment from Mirrool Creek.

In completing the flood study, the following activities were undertaken:

• Collation of historical flood information for the study area;
• Consultation with the community to acquire additional historical flood information;
• Development of a RAFTS hydrological model to simulate catchment rainfall-runoff;
• Development of a TUFLOW 2D/1D hydrodynamic model to simulate flood behaviour in the catchment;
• Development of a TUFLOW GPU 2D catchment model for Mirrool Creek to assist in the assessment of the flood hydrology;
• Calibration of the developed models using the available flood data, primarily relating to the March 1989 and March 2012 events;
• Prediction of design flood conditions in the catchment and production of design flood mapping series.

Through the undertaking of the flood study it has been found that the Main Drain J catchment is well regulated by the Main Canal and upstream storage area of Myall Park. The flood flows generated within the urban areas of Yoogali and Hanwood are therefore restricted to runoff from the catchment area downstream of the Main Canal. Coupled with the provision of significant manmade drainage, this results in a limited conveyance of flood flows within the broader floodplain extent. Out-of-bank flooding is predominantly characterised by flood waters ponding behind raised floodplain features such as road and rail embankments.

The performance of the model in representing catchment flood behaviour was supported by observations during the March 1989 and March 2012 flood events. The conditions observed in the Main Drain J catchment for the March 2012 event are generally representative of the modelled design 1 in 100-year probability event. This represents a significant change from the previously adopted flood study results which typically showed design 100-year flooding much more severe than March 2012 conditions. Accordingly, some changes may be anticipated to currently adopted flood risk and hydraulic category zones through the ongoing floodplain risk management process. Given the scale and nature of flooding, it is considered that suitable mitigation measures can be identified to address the existing flood risk to established urban areas in both Yoogali and Hanwood.

There is a reasonable level of uncertainty regarding the representation of embankment crest elevations in the Main Drain J model. This can influence localised flooding and the overall extent of the out-of-bank floodplain inundation. However, given the nature of the catchment flooding this does not have significant implications for the determination of flood planning constraints such as the defined floodways.

The March 2012 event also saw significant flooding of Mirrool Creek, which overtopped the Northern Branch Canal and spilled into the Main Drain J catchment, causing extensive flooding in Yenda. A catchment model was constructed for Mirrool Creek to represent this behaviour and assist in establishing appropriate design flood conditions for this mechanism.

For Mirrool Creek there is limited data from which to calibrate the models aside from the March 2012 flood event. This event has therefore been an essential platform from which to build an understanding of the catchment flood behaviour and quantifying design flood conditions. The lack of suitable calibration events for Mirrool Creek results in a large amount of uncertainty for design flood flow estimations. The hydrological response of the Mirrool Creek catchment is complex, being heavily influenced by the high infiltration rates of the sandy soils and the significant attenuation through the Barellan floodplain area.

The small number of recorded flood events in the Mirrool Creek catchment also reduces the reliability of flood frequency analysis, where there are large uncertainties in both the estimation of historic flood flows and the plotting position of the flood frequency.

Nevertheless, an attempt has been made to present these levels of uncertainty and determine an appropriate estimation of design flood flows for Mirrool Creek. This analysis will provide a platform for the future assessment of potential flood mitigation measures for Yenda.

The observed flood conditions for Mirrool Creek for the March 2012 event are estimated to be in excess of the 1% AEP (1 in 100-year) design conditions. The flood risk to Yenda from Mirrool Creek floodwaters emanates as the EMR capacity is exceeded. With both siphon and flood gates fully operational, this flow capacity may be expected to be exceeded for events in excess of the 2% AEP (1 in 50-year probability event). The current decommissioned status of the EMR flood gates structures significantly reduces the capacity to transfer Mirrool Creek flood flows across the Canal to the order of a 5% AEP (1 in 50-year probability) design standard. Accordingly, substantial flood mitigation measures may be required to provide an increased flood immunity to the Yenda township.

This flood study forms the basis for the subsequent floodplain risk management activities, being the next stage of the floodplain risk management process. The Floodplain Risk Management Study will aim to derive an appropriate mix of management measures and strategies to effectively manage flood risk. The findings of the study will be incorporated in a Plan of recommended works and measures and program for implementation.